Rasch Py

Rasch Py is a Python library that helps you implement Rasch models in Python. The 'Model' objects can be employed to output the relevant parameters and visualizations. Alternatively, the individual functions can be used to on already existing Rasch parameter outputs.

Code overview

Import

from rasch import DichotomousRaschModel #the main model object
from rasch.utils import generate_sample #a sample generator to test our functions

Sample data creation

We can create an example dataset with generate_sample(). When used with external data, make sure that it is formatted in long format with a single row per person and a column per item.

num_people = 10  # Number of people (rows)
num_items = 10    # Number of items (columns)
df = generate_sample(num_people, num_items)
print(df)

Model implementation and parameter estimation

Once the model object is created, difficulty and ability parameters can be estimated.

model = DichotomousRaschModel(df) #Create the model object
items, people = model.estimate_parameters() #Estimate the model parameters

Rasch plots

After the parameters have been estimated, ICC, IIF, and TIF plots can be generated in a single line of code.